Diabetic retinopathy is a significant cause of vision loss. In an attempt to determine how renin-angiotensin system blockers might reduce the burden of diabetic retinopathy, international researchers designed two randomised double-blind placebo-controlled trials.1 In 309 centres worldwide, normotensive participants with and without retinopathy were randomly allocated to groups receiving candesartan or placebo. The aim was to ascertain if the angiotensin II receptor antagonist could reduce the incidence of diabetic retinopathy in diabetics without the complication or slow its progression in those who already had the diagnosis. The results in over 3000 patients followed for 4 years suggested a role for candesartan in reducing the incidence of retinopathy, but no apparent beneficial effect on the progression of the disease in early cases. However, an accompanying comment by two experts from Australia and Singapore2 discusses the findings and the post-hoc statistical analysis, which suggested that treatment could increase regression of retinopathy in patients with early disease.